Abstract: This study presents an improved experimental protocol to overcome the limitations observed in the traditional undergraduate teaching experiment “fluorescence analysis for determining ohydroxybenzoic acid and m-hydroxybenzoic acid mixtures”, which suffers from methodological imperfections, significant measurement errors, and inadequate data processing training. The enhanced approach combines mathematical modeling with MATLAB-based data processing for analyzing mixed systems of o-hydroxybenzoic acid (o-HBA) and m-hydroxybenzoic acid (m-HBA). Emission spectra were acquired at a fixed excitation wavelength, followed by spectral deconvolution using Gaussian fitting and least-squares peak deconvolution algorithms implemented in MATLAB. This method enables accurate quantitative analysis of each component while requiring only single-pH condition measurements, thereby eliminating pH-dependent systematic errors inherent in conventional approaches. The experimental design integrates mathematical modeling, spectral visualization, and MATLAB programming, effectively enhancing students’ data processing capabilities, scientific reasoning, and interdisciplinary application skills. Preliminary implementation involving 112 undergraduate students over two semesters has yielded promising results, demonstrating significant potential for widespread adoption and educational reform.

Key words: Fluorescence emission spectroscopy, Two-component analysis, Mathematical modeling, Gaussian fitting, Peak deconvolution algorithm